Bioanalytical testing and quantitation methods suffer from interference from contaminants that can decrease or increase sensitivity to various analytes disproportionately to their abundance in the sample. Liquid chromatography-mass spectrometry/mass spectrometry
(LC/MS-MS) is the preferred method for drug metabolism studies; however matrix effects can lead to significant analytical errors and should be investigated to ensure that precision, selectivity and sensitivity are not compromised. (Little, J. L. et al. (2006) J. Chromatog. 833, 219). In particular, phospholipids such as phosphatidylcholines interfere with analyte ionization in electrospray MS detection by reducing analyte sensitivity, commonly referred to as ion suppression or matrix effects. See Ahnoff, M. and Hagelin, H. “Matrix Effects in Electrospray Ionization: Characterization of Plasma Phospholipids as Suppressors/Enhancers of Ionization Efficiency,” presented at the American Society for Mass Spectrometry, 52nd Conference on Mass Spectrometry (2004)). Further compounding the problem is that differing lipid composition of different samples, such as blood plasmas from varying animal species, can change the response of the analyte and cause problems in quantitation. These matrix effects are further discussed by the following presentations by Bennett, et al., which report the divergent calibration curves and the retention time shifts that can result from phospholipid contamination of bioanalytical samples: “Managing Phospholipid-Based Matrix Effects in Bioanalysis,” www.tandemlabs.com/capabilities_publications.html (accessed Feb. 26, 2007); “A Source of Imprecision Resulting from Ionization Suppression from Strongly Retained Phospholipids and Dioctyl Phthalate,” presented at the American Society for Mass Spectrometry, 52nd Conference on Mass Spectrometry (2004)).
In addition, the presence of contaminants can result in incomplete solvent extraction and hence underreporting of analyte concentrations, or can build up on analytical instrumentation, destroying sensitivity or resulting in downtime while cleaning procedures are instituted. Contaminants such as phospholipids have a tendency to build up on a typical reverse phase HPLC columns during repeated analyses of precipitated plasma samples. Accumulated phospholipids can bleed off in subsequent injections, causing a drift in analyte sensitivity over the course of multiple injections. See Bennett and Liang, “Overcoming Matrix Effects Resulting from Biological Phospholipids Through Selective Extractions in Quantitative LC/MS/MS,” presented at the American Society for Mass Spectrometry, 52nd Conference on Mass Spectrometry (2004). Removing the phospholipids requires extensive solvent washing to regenerate a column to proper condition.
Various approaches have been utilized in an attempt to solve these problems. Current methods for the removal of phospholipids from bioanalytical samples including liquid/liquid extraction (LLE) and solid phase extraction (SPE) are complicated and require a good deal of method development and the potential for analyte losses. For example, use of stronger eluting strength solvents in SPE can paradoxically result in decreased sample detection due to matrix effects, presumably due to contamination of the sample with phospholipids. However, limiting the eluting solvent strength to avoid contamination of samples with phospholipids can result in incomplete recovery of less polar analytes. LLE approaches require excessive amounts of labor and time to remove contaminating phospholipids, such as performing extraction and separation steps, and drying down or freezing samples, in order to remove the contaminants. For example, Bonfiglio et al. discuss the ability of several common extraction procedures to remove endogenous plasma components that cause ion suppression in electrospray ionization tandem mass spectrometry. LLE using methyl-t-butyl ether, SPE with Oasis and Empore, and acetonitrile (ACN) protein precipitation sample preparation methods were compared. These researchers found that ACN protein precipitated samples showed the greatest amount of ion suppression while LLE extracts demonstrated the least. In addition, the ion suppression was found to be analyte dependent, and associated with the most polar analyte. The least ion suppression for all analytes was observed in samples treated with both LLE and SPE. The authors conclude that there were most likely multiple endogenous components involved in ion suppression, and that the effects may persist well after the injection into the HPLC system is made, resulting in the collection of invalid data. A further filtration step was suggested in an effort to provide yet cleaner samples for analysis. (Bonfiglio, R., et al. (1999) Rapid Comm. Mass. Spectr. 13, 1175). However, use of multiple sample preparation steps is labor and time intensive, and increases the cost of performing analyses.
More recently, new approaches for removing phospholipids from samples have been attempted. For example, Johanson reported that use of strong cation exchange column to remove cationic lipids including phosphatidylcholines from lipid extracts resulted in the ready detection of peaks that had been completely suppressed in the crude extract. (Johanson, R. A., et al. (2007) Anal. Biochem. 362, 155). U.S. Patent Application Publication No. 20050054077 (Bennett, et al.) describes devices and methods for removing phospholipids from biological samples involving the use of a phospholipotropic multivalent cation coupled to a support. Such cations reportedly include transition metals, lanthanides or actinides, preferably cerium. Use of these phospholipotropic multivalent cation sorbents was further described by Van Home, et al., describing the sorbents as possessing high oxophilicity for the phosphate groups on the phospholipid molecules. (Van Home, K. C., et al. “Preventing Matrix Effects By Using New Sorbents to Remove Phospholipids from Biological Samples” (2003) presented at the Proceedings of the American Association of Pharmaceutical Scientists Conference). The authors reported a goal of providing facile removal of phospholipids from biological samples and extracts utilizing an extraction chemistry that would not remove desirable pharmaceutical analytes. Many different mechanisms reportedly were evaluated, including reverse-phase (nonpolar) and both anion and cation exchange. Phospholipid extraction was implemented via extraction sorbents used alone or in combination with protein precipitation, liquid liquid extraction (LLE) or solid phase extraction (SPE). In particular, the phospholipid content of extracts reconstituted in methanol was reportedly reduced by as much as 94-96% using the lanthanide sorbent alone. Use of a lanthanide extraction sorbent to remove phospholipid from a protein precipitated sample reconstituted in methanol reportedly resulted in phospholipid removal of about 92%-98% and enhanced detection of spiked analytes. However, the procedure required centrifugation followed by solvent evaporation and reconstitution, which is labor and time consuming and adds to the costs of performing analysis. Use of a lanthanide extraction sorbent to remove phospholipid from a sample after methyl-t-butyl ether LLE reportedly resulted in phospholipid removal of about 95%-97% but without significant enhanced detection of spiked analytes. The authors concluded that immobilized lanthanide metal centers are an essential element for highly selective binding for phospholipid extraction via binding to the phosphate groups.
Shen, et al. describe an evaluation of three different types of ion-exchange solid phase extraction media in an effort to determine the abilities of the media to remove phospholipids from analyte solutions. These authors reported that mixed mode phases fulfill the requirement of retaining both analytes and diverse metabolites, while reverse-phase retention mechanisms were detrimental in eliminating ion suppression caused by late eluting phospholipids, and advised using an ion exchange mechanism alone rather than mixed mode extraction phases. (Shen, J. X., et al. (2005) J. Pharm. Biomed. Anal. 37, 359).
U.S. Pat. No. 5,885,921 to Krupey describes the use of hydrophobic silica adsorbents for the removal of lipids in samples, sold under the brand name Cleanascite™ for lipid adsorbent and clarification agent for pretreatment of samples prior to further purification. The adsorbent is added to samples and then the sample is centrifuged to remove the adsorbent containing bound impurities. However, this procedure requires two steps to add the sorbent and then remove it, and risks removal of analytes from samples.
U.S. Pat. No. 5,759,549 to Hiltunen describes the use of supercritical fluid extraction for the isolation of lipids from mixtures of lipids. However, this procedure requires specialized equipment and adds to the expense and cost of removing lipids from samples.
Little et al. describe an “in source multiple reaction monitoring” method for monitoring method development of pharmaceutical analysis in an effort to determine whether there are co-eluting matrix constituents resulting in ion suppression of analytes of interest. However, these procedures do not remove the matrix constituents causing the ion suppression, but merely attempt to work around the problem, and require the elution of the most hydrophobic lipids from the column after each analysis. (Little, J. L. et al. (2006) J. Chromatog. 833, 219).
Therefore, numerous methods for removing lipid contaminants in biological samples are available, although the procedures are time and labor intensive. Further, samples also contain contaminating proteins, which must also be removed. However, use of denaturing solvents to effect precipitation results in greater extraction of lipid contaminants and significant ion suppression of analytes present. Therefore, there is a minimum of two steps needed to prepare a bioanalytical sample for analysis, if the researcher hopes to maintain his equipment in good working order, and to achieve reliable and accurate quantitation of bioanalytes. When working with small sample volumes, or when multiple testing is needed, such multiple sample preparation procedures are likely to reduce sample such that an insufficient amount remains for the testing required. In addition, multiple sample preparation steps result in a loss of researcher time and labor and raise the costs for the analytical testing laboratory.
Accordingly, there remains a need for rapid procedures that can remove both phospholipids and other agents causing matrix effects and proteins from a bioanalytical sample prior to performance of analytical procedures.